INDEX

Explanations

instances where something is admitted or acknowledged

oai_token-act-pair · gpt-3.5-turbo

expressions related to the concept of admitting or acknowledging something

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

rior

-0.89

chn

-0.77

nton

-0.73

chnology

-0.67

miah

-0.65

lets

-0.65

gm

-0.64

locked

-0.63

acements

-0.61

otor

-0.61

POSITIVE LOGITS

ibility

0.98

IBLE

0.87

 wrongdoing

0.85

iary

0.82

 defeat

0.79

 admit

0.75

essors

0.75

elsen

0.75

aneers

0.75

ibilities

0.74

Activations Density 0.022%

instances where something is admitted or acknowledged

expressions related to the concept of admitting or acknowledging something

No Comments

No Known Activations